The case against this Regulation

Drafted first, before any article, on the discipline this project inherits: if the instrument cannot survive this file, we fix the instrument, not the prose. Each objection is stated in its strongest form, with its best sources. Each ends with the design consequence it imposes. The consequences accumulate into the constraints table at the end, which is the checklist the articles must clear before anything else is written.

Objections marked CONCEDED are limits of the instrument that the memorandum states plainly rather than argues away. That is not a courtesy. It is what distinguishes a legal proposal from campaign copy, and it is the entire credibility strategy of this project.

---

A. Legal

1. This is expropriation

The objection, at full strength. A statutory requirement that undertakings issue equity warrants to a public reserve takes property. Article 17 of the Charter of Fundamental Rights protects the right to own, use and dispose of lawfully acquired possessions, and permits deprivation only in the public interest, in cases and under conditions provided for by law, and subject to fair compensation paid in good time. A compulsory warrant dilutes existing shareholders with no compensation at all; the dilution is the point. Article 345 TFEU adds that the Treaties shall in no way prejudice the rules in Member States governing the system of property ownership. The Court has narrowed Article 345 considerably, but a hostile Legal Service reading has ample material, and shareholders of affected firms will litigate from day one.

What it gets right. A warrant requirement does transfer value from existing shareholders to the reserve. Pretending dilution is not a taking of value would be dishonest, and the memorandum must not do it.

The answer the instrument must give. Three structural choices, none optional. First, the warrant must be drafted as a condition of access to the Single Market for a defined category of undertaking, prospective and uniform, in the family of regulatory obligations the Court has upheld when proportionate (capital requirements, universal service obligations, DMA gatekeeper duties), not as a seizure of existing holdings. Second, it must apply only above high, objective thresholds, so proportionality review has something to hold on to. Third, it must carry consideration: the reserve is a passive, non-controlling holder, the warrant crystallises at the statutory liquidity, extraction or elapsed-time triggers, and the covered undertaking receives the legal certainty of a single harmonised regime in place of twenty-seven national experiments. Whether that consideration is sufficient is the single largest legal risk in the project, and Gate 1 exists to test it.

Two authorities belong in this file because a hostile reader will bring them, and because each is less one-sided than it first appears. The first is Strasbourg's rule that a deprivation without an amount reasonably related to the value taken is normally disproportionate (James and Others v United Kingdom, 1986). The same judgment holds, in the same breath, that legitimate objectives of public interest, and expressly measures of economic reform or measures designed to achieve greater social justice, may warrant reimbursement at less than full market value. That is the authority under which this instrument has to be argued, and it is available on its terms rather than in spite of them: a 3 % dilution crystallising at a statutory event, calibrated to a decoupling of output from labour, is a measure of economic reform if anything is.

The second is Article 345 TFEU. It provides that the Treaties shall in no way prejudice the rules in Member States governing the system of property ownership, and the hostile reading is that a mandatory transfer of 3 % of private enterprise value is a partial socialisation the Union may not decree. That reading mistakes the addressee. Article 345 preserves the Member States' freedom to choose between public and private ownership against the Treaties; it is not a charter of immunity for property against Union legislation, which is what Article 17 of the Charter governs and what this memorandum answers under Article 52(1). The Court has read Article 345 narrowly in precisely that direction, holding that a property regime chosen by a Member State remains subject to the fundamental freedoms (Essent, C-105/12 to C-107/12). The instrument leaves every Member State's system of ownership exactly as it found it: the shares subscribed are ordinary shares under national company law, held by an owner with no special rights, and no national rule about who may own what is displaced.

The drafting research (18 August) settled the architecture to use: BRRD is the judicially validated template for interfering with equity by act of law (Kotnik C-526/14, Ledra C-8/15 P, Aeris Invest C-535/22 P), and the instrument adopts its machinery: an honest interference recital naming Charter Article 17, a full proportionality recital under Article 52(1) of the Charter, and a quantified executional safeguard with independent, separately challengeable valuation. One adaptation is mandatory: BRRD's counterfactual is insolvency, and a permanent regime cannot lean on crisis reasoning (Dowling C-41/15). The distinction has to be stated rather than blurred: bank resolution dilutes shareholders whose shares would be worth nothing in the counterfactual insolvency, and these undertakings are healthy going concerns whose shares are worth a great deal. BRRD is therefore borrowed for its machinery, not for its justification, and this file does not pretend otherwise; the justification is the one in James and in Hauer, and it stands or falls on proportionality rather than on a crisis that is not happening.

One further authority is worth stating precisely, because stated loosely it would be worth nothing. In Sky Österreich (C-283/11, 22 January 2013) the Grand Chamber upheld a Union measure requiring holders of exclusive broadcasting rights to grant competitors access for short news reports, with compensation capped at the additional costs directly incurred, which is below what the market would pay. The rights-holders were solvent, there was no crisis and no counterfactual insolvency. That case was decided under Article 16 of the Charter, the freedom to conduct a business, and not under Article 17. It is not authority for a taking of equity, and this file does not offer it as one; hostile counsel would be entitled to dismantle any such use of it, and would enjoy doing so.

What it does carry is narrower and still useful. This instrument interferes with Article 16 as well as Article 17, because compelling an undertaking to issue shares it did not choose to issue is an interference with the conduct of its business, and the Commission's Legal Service will see both. On that limb Sky Österreich is directly in point, and it says three things: the Article 16 freedom is not absolute but must be viewed in relation to its social function; the Union legislature may set consideration below market value where the public interest requires it; and the test is the ordinary Article 52(1) one. The older line supports the same reading of business interests, which enjoy no protection as mere commercial expectations of future profit (Nold 4/73). Booker Aquaculture (C-20/00 and C-64/00) is sometimes cited alongside, and this file does not lean on it: it concerned the destruction of diseased stock under an animal-health regime, which is a public-emergency justification this instrument cannot claim.

The Article 17 limb therefore still rests where it rested, on James and on Hauer, and the honest statement is that no decided case puts a permanent, non-crisis, uncompensated equity dilution of a healthy undertaking on the right side of Article 17. That is the largest legal risk in the project and it is stated as such in the answer above. So our floor is executional rather than counterfactual: the interference can never exceed the stated 3 % in execution, its price is set by the liquidity event itself under an independent and separately challengeable valuation, and no application of the instrument may take more than the interference it names. The doctrine underneath is older than any of this: property in the Union legal order is not an unfettered prerogative but is protected in its social function, and may be restricted in the general interest where the restriction is proportionate and leaves the right's substance intact (Hauer 44/79; Bosphorus C-84/95). The instrument's restriction is quantified, event-bound and substance-preserving by construction, which is what those cases require the legislature to show.

Design consequence. DC-1: prospective warrant on future value creation at defined events, never retroactive transfer of existing shares. DC-2: high group-consolidated thresholds. DC-3: passivity, and crystallisation only at defined statutory events, never at a discretionary or political one, written into the instrument itself.

2. This is a tax, and the EU may not levy it this way

The objection, at full strength. Call it a warrant; it functions as a levy. Article 114(2) TFEU excludes fiscal provisions from internal market harmonisation, so if the measure is fiscal in substance the chosen legal base collapses; the honest bases would be Article 113 or 115, which require unanimity in Council, which is unobtainable. The Commission has refused ECIs that drift into own-resources territory, and the Legal Service reads substance, not labels. The dividend side makes it worse: a recurring payment to every adult, funded by an obligation on firms, looks like a tax-and-transfer scheme wearing a corporate-finance costume.

What it gets right. The boundary is real and the characterisation battle decides registrability. This is the objection most likely to kill the full ask at Gate 1.

The answer the instrument must give. The measure must take nothing in money from any undertaking in any year. No cash flows from firms to the state at all. The reserve receives instruments, holds them, and distributes returns on what it owns, exactly as any shareholder does. Dividends to citizens are property income from an owned portfolio, not the proceeds of a levy: the Norwegian fund's distributions are not a tax on anyone. The drafting must police this line everywhere: no revenue-based charges, no minimum payments, no cash-settlement options that would let the obligation collapse into a fee. And the ask must be layered so that if the Legal Service still reads it as fiscal, the severable outer layer (assess and propose instruments for citizen participation in automated productivity gains) registers on its own.

Recalibration from the drafting research: registration is a lower hurdle than this objection assumes. The test is "manifestly outside" the Commission's powers (Reg 2019/788 Art 6(3)(c)), partial registration is judicially established (C-899/19 P), and the Commission registered the EU wealth-tax ECI in 2023. The characterisation fight is real but is fought in Council, after registration. The layering therefore protects the campaign moment; the warrant-not-levy drafting protects the instrument's life after it.

Design consequence. DC-4: no monetary flow from undertakings; instruments only. DC-5: distributions defined as property income of the reserve. DC-6: severable layering for partial registration.

3. The Union has no competence to pay citizens a dividend

The objection, at full strength. Even if the warrant survives, the dividend side has no home. The Union budget operates under an own-resources ceiling; a Union body paying a recurring universal benefit to every adult has no Treaty basis; social security design is a Member State competence; and subsidiarity review would ask, fairly, why citizen accounts must be European at all when Ireland, Denmark and the Netherlands run national systems that work. Table 29 makes the point against us: pension funding runs from 49.1% in the Netherlands to 0.0% in France. Systems this different cannot be harmonised, and should not be.

What it gets right. The Union genuinely cannot and should not run twenty-seven citizens' accounts from Brussels, and the proposal dies at subsidiarity review if it tries.

The answer the instrument must give. Split the instrument along the competence line. The Union harmonises what is genuinely single-market: which undertakings issue warrants, on what terms, to what kind of reserve, with what governance. Custody and distribution federate to Member States through existing rails: Denmark has LD, Ireland has MyFutureFund, Poland has PPK, the Netherlands is mid-conversion into individual DC pots. The precedent is PEPP: a Union framework, national compartments. The Union never touches the money; it defines the instrument and the minimum standards (universality, lock-up, raid-proofing) that national implementations must meet.

Design consequence. DC-7: Union-level warrant and reserve standards; Member State custody and payout through existing pension rails. DC-8: minimum standards, not uniform machinery.

---

B. Economic

4. Firms will pass the cost to consumers, so citizens pay their own dividend

The objection, at full strength. The incidence literature on corporate taxation is unambiguous that legal and economic incidence differ; a substantial share of corporate burdens lands on workers and consumers. A warrant obligation raises the cost of operating in Europe; covered firms reprice; the citizen's dividend arrives net of the citizen's own higher prices. The scheme is then a circular pump with deadweight loss.

What it gets right. Some pass-through of any burden is real and claiming zero incidence would be amateurish.

The answer the instrument must give. Equity dilution has materially different incidence from a flow charge. A levy on revenue enters marginal cost and prices directly; a one-time issuance of warrants exercisable at future liquidity events changes the division of a future capital gain among shareholders and does not enter this year's marginal cost at all. The firm's optimal price today is unchanged by who owns claims on its eventual exit value. Pass-through is not zero, because expected dilution can raise the cost of capital at the margin, and the memorandum should say so, with the honest note that this effect is second-order next to any revenue levy, which is precisely why the instrument is a warrant and not a levy.

Design consequence. DC-9 (reinforces DC-4): the obligation must never be convertible into a flow charge, because the incidence answer depends on it.

5. Covered firms will leave, or never come

The objection, at full strength. Draghi's report already concedes the ground: it is too late for the EU to develop systematic challengers to the major US cloud providers; the US ITK market grows at 12.7% against Germany's 4.1%; only four of the world's top fifty technology companies are European. Add a warrant obligation and the marginal AI investment goes to London, Zurich or Austin. Worse, thresholds invite structuring: a revenue-per-employee test is gamed by pushing headcount into subcontractors, exactly the offshore structuring Capgemini's numbers already show at scale.

What it gets right. Threshold gaming is certain, not possible, and the competitiveness anxiety is the strongest political headwind in Brussels this decade.

The answer the instrument must give. The obligation attaches to selling into the Single Market, not to being located in it, exactly as the DMA and GDPR attach. The empirical record since is that gatekeepers absorbed designation and stayed, because 450 million high-income consumers are not optional; the DMA investigations into AWS and Azure did not produce an exit, they produced compliance teams. Structuring is answered by consolidation: thresholds computed on group-consolidated figures including contracted-out labour by economic substance, with the burden on the undertaking to show otherwise. And the honest concession: at the margin some investment will route elsewhere, which is a real cost, to be weighed in the memorandum against the documented cost of the alternative, which is that the gains concentrate entirely outside Europe anyway. The Synergy history is the exhibit: Europe declined to regulate its cloud market into openness, and its providers fell from 29% to 15% of their own home market unregulated.

Design consequence. DC-10: market-access nexus, not establishment nexus. DC-11: group consolidation with substance-over-form headcount rules.

6. Warrants on private companies cannot be valued, voted or sold

The objection, at full strength. The covered class is dominated by private undertakings. A reserve holding warrants on private micro-giants holds paper with no market price, no liquidity and no exit; either it pressures for early listings, distorting capital markets, or it sits on unvalued claims for a decade and the dividend it promises cannot be paid. Meanwhile the governance question is a trap in both directions: a passive mega-holder is the feeble owner of the Bebchuk critique, and an active one is a political sovereign shareholder in every major firm. Greece's HCAP shows the overcorrection: maximal raid-proofing achieved by placing the asset beyond citizens' reach entirely.

What it gets right. All of it. This is the hardest design problem in the instrument, harder than the legal base.

The answer the instrument must give. The warrant is dormant until crystallisation, whether at a liquidity event (listing, change of control, or qualifying secondary sale) or, under Article 5(3), where shareholder extraction exceeds the stated share of covered turnover or seven years have elapsed since issuance, whichever occurs first. Until then it requires no valuation, pays nothing and votes nothing; at the event it converts at the event price, with no discretion. This matches the book's own window (claim the stake while the asset forms, crystallise when the market prices it) and removes the valuation and governance problems in one move: the reserve holds non-voting economic interests, permanently, by statute, accepting the Bebchuk cost deliberately because the alternative, a politically voted stake in every large firm, is worse. The dividend in early years is funded by crystallisations, not by holdings, and the memorandum must say plainly that the dividend starts small and compounds, Norway-style, and that anyone promising otherwise is not us. Two-thirds of Norway's fund is compound return; the honest pitch is the rule, not the first cheque.

One refinement has been considered and rejected. It is put that valuations of fast-growing private undertakings are volatile enough that disputes over the dilution ratio could bog the instrument down in litigation, and that an expedited binding arbitration should therefore sit between the valuer and the courts. The instrument already answers the problem the proposal aims at, and the proposal would cost more than it saves. Article 6(5) provides that a challenge suspends neither the liquidity event nor the subscription, and Article 6(4) corrects an erroneous valuation afterwards in either direction, so a dispute delays nobody: this is the resolution-law pattern, where litigation runs beside the transaction rather than across it. Where the event itself sets a price, Article 6(2) makes the valuation arithmetic on that price rather than an opinion about it, which is where volatility would otherwise enter. And an arbitral tier whose award bound the parties would meet Article 47 of the Charter on access to a court, and the limits on conferring discretionary judgement on bodies of the Union; a tier whose award did not bind them would add a stage without removing one.

Design consequence. DC-12: event-triggered crystallisation, no ongoing valuation. DC-13: permanently non-voting economic interests. DC-14: the dividend is communicated as compounding from small, never as immediate income. DC-34: no tier may be inserted between the valuation and the courts; correction is ex post and the transaction never waits.

---

C. Empirical

7. The European labour share has not fallen, so the diagnosis is wrong

The objection, at full strength. AMECO, verified at source: the EU27 adjusted wage share fell 1.2 points in thirty years; Germany's is at a series high; France is above its 1995 level; compensation of employees was a larger share of EU output in 2025 than in 2015, 2019 or 2022. The corporate profit share spike of 2022 has fully reversed to below its 2019 level. The Regulation's premise, that machine-driven gains are leaving European labour, is contradicted by the best available aggregate data, and any recital claiming otherwise is checkable and wrong.

What it gets right. Everything it states. This project's own evidence base established it, and no recital may argue a European labour-share collapse.

The answer the instrument must give. The premise is ownership, not the wage share. Sixty per cent of euro-area households own their home; eleven per cent own listed shares; the top decile holds eighty-three per cent of directly held shares against the bottom half's two; the bottom half's portfolio is sixty-three per cent housing and three per cent equity. That distribution is current, verified and undisputed, and it means that however large the machine's dividend turns out to be, almost no European holds an instrument that pays it. The Regulation is insurance whose premium is cheapest before the event: if the mechanism documented at the platform layer (French software and cloud growing at +8.2% on price increases while the services layer shrinks; German software at +9.9% against services at +3.1%) reaches the aggregate labour market, the stake exists; if it never does, citizens own a diversified claim on European technology, which is not an injury. The in-time test is the answer to the premature-legislation charge, not a vulnerability of it.

Design consequence. DC-15: recitals argue ownership concentration and mechanism, never wage-share decline. DC-16: the memorandum presents the instrument's value under BOTH futures, arrival and non-arrival.

8. The best microdata finds nothing for AI to answer for

The objection, at full strength. Humlum and Vestergaard, on Danish registers covering twenty-five thousand workers: precise null effects on earnings and hours, nothing above two per cent, two years after ChatGPT. The OECD finds no break in postings for exposed occupations and a flat euro-area youth gap. The German federal government told the Bundestag there are keine Hinweise that AI has reduced entry-level chances. Only a fifth of EU firms with ten or more staff used any AI in 2025. Legislating a permanent constitutional-grade structure on this evidence is panic dressed as foresight.

What it gets right. The nulls are real, well-identified and from the best registers in Europe. The memorandum cites them in full or loses its credibility.

The answer the instrument must give. Three things, honestly. First, the nulls measure wages and hours, not ownership: a uniform shift of returns from labour to capital is structurally invisible to difference-in-differences designs, which the Danish authors themselves concede (the absence of measurable labour market effects is not evidence that nothing is happening), and their working paper's estimate that only three to seven per cent of productivity gains passed through to earnings vanished from the published version along with the paper's own title. Second, the diffusion number cuts both ways: a mechanism used by a fifth of firms has not yet had its aggregate chance, which is the argument for building the instrument before rather than after. Third, the memorandum should carry its own falsification condition, in the open: if adoption passes a quarter of firms and neither factor shares nor ownership-weighted gains have moved by a defined date, the accelerationist premise is wrong and the case reduces to the ownership-gap argument alone, which stands on its own feet. A proposal that states what would refute it is a different genre from everything else in Brussels, and that is the point of this project.

Design consequence. DC-17: cite the strongest nulls in the memorandum itself. DC-18: a stated falsification condition with a date.

---

D. Political

9. You are building the next honeypot, and Europe raids honeypots

The objection, at full strength. In one article of law, Poland cancelled 51.5% of the units in every member's pension account and took the Treasury bonds first. Spain drew its reserve fund down 97% and passed the protection law afterwards. Hungary took the lot. Ireland's NPRF, the closest thing to this proposal an EU state has built, was liquidated into a bank rescue within a decade of its creation. Even the Union itself announced InvestAI at 200 billion and reprogrammed existing funds to stage it. The historical base rate for European states leaving large pools of citizens' capital alone is poor, and a bigger pool is a bigger prize. Estonia adds the mirror case: given the claim and the exit simultaneously, a quarter of the fund walked out in a month, and the leavers earned below the median.

What it gets right. This is the book's own chapter 10 thesis returned as an objection, and it is the strongest political argument in the file. The raid is not a tail risk; on the European record it is the modal outcome.

The answer the instrument must give. Design against the actual statutes, clause by clause. Against Poland: the reserve may not hold the sovereign debt of any Member State or of the Union, removing the self-dealing channel that made OFE worth cancelling. Against Spain: the prohibition on disposal is in the founding Regulation from day one, not retrofitted, and amendment of the protection provisions is reserved to express legislative change accompanied by a published independent assessment of the effect on holders; a Regulation cannot prescribe voting thresholds or delays for future legislatures, and pretending otherwise would hand critics the easiest ridicule in the file. Against Ireland: no emergency-use clause of any kind, because the NPRF's raid was lawful under its own emergency provisions. Against Estonia: the periodic entitlement is incapable of surrender or redemption against payment, so there is no exit for a buyout to purchase, which is the failure Estonia proved; amounts already distributed are the holder's property, inheritable, the Danish device that held. Against Greece: raid-proofing must never be achieved by removing the citizen's claim, which is the failure dressed as success. And one honest sentence the memorandum must contain: no drafting defeats a determined future sovereign; the design goal is to make the raid loud, slow and electorally expensive, which is the most any constitution has ever achieved.

Design consequence. DC-19: no sovereign debt holdings. DC-20: entrenchment as friction from day one: express amendment only, a published independent assessment, and honesty about Treaty limits. DC-21: no emergency clause. DC-22: individual claims incapable of surrender; distributed amounts owned and inheritable. DC-23: the raid-resistance claim is stated as friction, never as impossibility.

10. The EU cannot even disburse what it announces

The objection, at full strength. The AI Gigafactories call opened eighteen months after the 200 billion announcement; construction is promised from 2027; the Commission's sovereign-cloud tender awarded 180 million against a single hyperscaler's 33.7 billion in Spain. The institutional metabolism that would operate this Regulation moves at a pace the underlying economy does not recognise. A citizens' reserve run at that pace would crystallise warrants years late and distribute nothing for a decade.

What it gets right. The disbursement record is accurately damning and the memorandum gains nothing by disputing it.

The answer the instrument must give. Put no Union spending machinery on the critical path. The obligation runs directly from covered undertakings to the reserve: warrants issue by operation of law upon crossing the thresholds, crystallise by operation of law at events, and the reserve's task is custody and distribution, not procurement. Nothing needs to be built, tendered or disbursed for the instrument to function; the contrast with InvestAI is the design, not an embarrassment to it.

Design consequence. DC-24: self-executing obligations by operation of law; no grant, tender or programme machinery anywhere in the instrument.

---

E. Philosophical

11. Everyone gets richer anyway: the consumer surplus answer

The objection, at full strength. Nordhaus estimated that innovators capture around two per cent of the social value of their innovations; the rest diffuses to consumers through falling prices and new possibilities. The steam engine made everyone richer without a single citizen holding a warrant on Boulton and Watt. If AI follows the pattern, the ownership question is a distraction: the gains arrive in everyone's basket regardless of whose name is on the equity.

What it gets right. Consumer surplus is real, large and the main channel through which technology has historically raised living standards. The book concedes it; the memorandum must too.

The answer the instrument must give. Two facts sit beside the diffusion story without contradicting it. First, the lag: British real wages stagnated for roughly half a century after Watt while output exploded, and the broadening arrived through unions, franchise and factory legislation, not through prices alone; the people who lived inside the pause did not get those decades back, and whoever owned the mine did not wait. An instrument claimed at the door is the difference between experiencing the pause with and without an asset. Second, the stock and the flow are different questions: cheaper consumption and concentrated wealth are compatible, and the euro area currently demonstrates the combination, with record employment beside an ownership distribution of eighty-three against two. The dividend does not obstruct diffusion; it adds an ownership channel beside the price channel.

Design consequence. DC-25: the memorandum affirms consumer surplus and positions the instrument as additive to it, never as a correction of a falsehood.

12. A dividend buys neither status nor purpose

The objection, at full strength. Positional goods reprice against any universal transfer: give everyone more and the house in the right street costs more. And income without work solves rent, not Tuesday afternoon: the structure, standing and belonging that employment supplies as a by-product are not in the reserve's gift. The proposal oversells what capital income can do for a displaced person.

What it gets right. CONCEDED in full. These are the book's own chapters 11 and 13 and they bind its regulation exactly as they bind its argument.

What the memorandum must therefore say. The instrument's scope statement, prominently: this Regulation addresses the distribution of machine-generated capital income and nothing else. It does not promise status, purpose, meaning or the best table, and any campaign material that implies otherwise is wrong by the proposal's own text.

Design consequence. DC-26: an explicit scope-and-limits recital.

13. This is universal basic income with extra steps

The objection, at full strength. Strip the corporate finance and a citizen receives a recurring unconditional payment from a public body. That is UBI, a policy the electorate has weighed repeatedly and cheaply funded versions of which already failed to collect even an ECI's signatures. The warrant machinery is complexity added to disguise a transfer as a return.

What it gets right. The payment is indeed unconditional and universal, and the family resemblance is real at the point of receipt.

The answer the instrument must give. The difference is not in the receiving but in the standing. A benefit is a flow from the annual budget, renegotiated annually, cancellable by simple majority, funded by a shrinking wage base; the account here is property, in the citizen's name, inheritable, funded by returns on assets the citizen collectively owns. Poland could cancel account units by statute precisely because they held claims on the state itself; Denmark's LD has paid for forty-six years because it holds market assets in named accounts. The tagline is the answer in six words: capital for all, so the dividend follows. The order is the argument.

Design consequence. DC-27 (reinforces DC-22): named, inheritable individual accounts, so the UBI comparison fails on legal form, not on rhetoric.

14. Pensions already do this; use them

The objection, at full strength. Europe already possesses the machinery for broad capital ownership: the Netherlands has just moved 605 billion into individual DC entitlements, Sweden's premium pension has compounded at 5.35% real since 2000, and the American case shows retirement accounts spreading equity to 58% of households without any statutory warrant. Build pensions, not novelties.

What it gets right. The delivery rails exist, work and are trusted, and any design that ignores them is wasteful.

The answer the instrument must give. Pensions convert wages into ownership, and the wage is the input this transition erodes; every pension vehicle in Europe assumes an employer and a payroll deduction, which is exactly the assumption chapter 6 shows failing. The funded share of accrued pension wealth is 7.8% in Belgium and 0.0% in France: the rails reach the employed of the funded countries and no one else. The warrant severs the funding source from the wage: the reserve's returns flow into precisely those national account rails (DC-7) whether or not the citizen ever held a payroll job. Pensions are the pipe; this is a new source feeding it.

Design consequence. DC-28 (reinforces DC-7): distribution through existing national pension rails; the novelty is confined to the funding source, where it is necessary.

---

15. A euro a year is an insult, not a policy

The objection, at full strength. Run the instrument's own simulator on cautious assumptions and it pays a citizen a euro or two a year for its first decade. No rational person values that; no voter campaigns for it; no journalist resists the headline. The apparatus is grotesquely disproportionate to its payload: a new Union body, a designation regime, valuation machinery, comitology, penalties reaching 10 % of worldwide turnover, all to deliver less than the price of a coffee. Worse, the project's own honesty doctrine forbids promising more. An initiative that must, by its own rules, tell every signer "you will not feel this for twenty years" has chosen a message no mass campaign has ever won with. Basic income at least promises the rent.

What it gets right. The early flow is genuinely small, and the campaign is structurally barred from inflating it. The collection risk is real: deferred rewards lose to immediate ones on every doorstep.

The answer the instrument must give. Three parts. First, small cash distributions do not by themselves show that the automation economy is small or that the premise has failed. A large ownership stake can produce little realised income. Article 14's specified indicators and review process test the premise; dividend size alone does not. The generational capital objective and the availability of cash income need to be assessed separately. Second, the claim can only be bought early. The Danish worker's frozen 1978 instalment of DKK 4 368, pointless money at the time, is DKK 119 506 today; two thirds of Norway's fund is compound return, not oil. The alternative timing, claiming the stake after the gains are visible and the owners entrenched, is expropriation and politically impossible. The euro buys the certificate, and the certificate is the point. Third, the stock outruns the flow: on the same cautious assumptions the Reserve stands at roughly EUR 400 of owned capital behind every citizen by year thirty, before any single year's payout impresses. A campaign that shows the payout without the stake is misdescribing its own instrument.

Design consequence. This objection is why DC-14 exists (never lead with an early-year figure), why Article 14(3) carries the falsification condition on its face, and why Annex II distributes income and never principal. It adds two mechanical rules of its own: wherever a per-citizen payout figure is shown, the per-citizen stake in the Reserve is shown beside it (DC-31); and where a year's distributable amount would be too small to be worth transferring, the interval lengthens rather than the money vanishing, subject to a hard backstop of one distribution in every three years and a duty to publish what was not paid (DC-40). The threshold is a ratio to the cost of making the payment, not a figure: a number in an annex ages, and hands a headline to anyone who wants one.

16. This is a golden share, and the Court strikes golden shares down

The objection, at full strength. For twenty years the Court has dismantled every special public equity position in the internal market. Commission v Germany (C-112/05) struck the Volkswagen Law because voting caps and a blocking minority gave public authorities influence exceeding their investment and thereby deterred direct investment; the Portuguese, French, Italian, British and Dutch golden shares fell the same way under what is now Article 63 TFEU. A Union-chartered Reserve holding a statutory 3 % of every covered undertaking is a golden share cast as a regulation: a permanent state-adjacent shareholder no investor chose, planted in the capital structure of every frontier firm, deterring exactly the cross-border investment Article 63 protects. Article 345's neutrality about systems of ownership did not save the Volkswagen Law and will not save this. The deterrence is not hypothetical: every venture round in a covered undertaking now prices a mandatory future dilution.

What it gets right. The dilution is real and investors will price it, and any position that carried control-flavoured rights would fall exactly as Volkswagen fell. The golden-share cases are the controlling jurisprudence for any public equity position, and the instrument must be drafted against them, not around them.

The answer the instrument must give. The golden-share line condemns one thing: special rights of control disproportionate to investment, voting caps, blocking minorities, approval vetoes, board seats, the machinery by which a state steers a company it does not own. The instrument constructs the exact inverse, in the articles rather than in assurances. The Reserve's holding carries no vote, ever (Articles 5(4)(a) and 9(1)(a)); no board presence (Article 9(1)(b)); no instructions (Article 9(1)(c)); no acquisitions beyond the warrant and index-style diversification (Article 9(1)(d)); no leverage or derivatives that would make it a strategic actor (Article 9(1)(e) to (g)). What remains is pure economic participation, the position of any passive minority holder, which is the position Norway's fund holds at comparable scale across European listed undertakings without an Article 63 case ever being brought. What the case law demands of any restriction that survives, the instrument answers on its face: non-discrimination (identical treatment of Union and third-country undertakings under Article 3), an overriding general interest stated in the recitals, and proportionality carried by the fixed 3 %, the independent and separately justiciable valuation (Articles 5(9), 6 and 7) and the Article 52(1) balance. And unlike every struck golden share, this is not a Member State reserving national influence against integration: it is a uniform Union rule for the whole internal market, and its uniformity removes the divergence that national participation schemes would create. The honest residue is that mandatory future dilution is itself a cost investors will price; objection 4 prices it, and proportionality, not denial, is the defence.

Design consequence. DC-13's permanent non-voting rule and Article 9's conduct prohibitions are this objection's answer in law. What they cannot do is bind a future legislature, and this memorandum does not pretend otherwise: the drafting of Article 12(3) records that binding future legislatures failed the gates in an earlier round, and an attempt on 21 August 2026 to give the rule operative force in that paragraph was abandoned because hostile counsel preferred it present, reading it as a Union-law command that the Reserve be stripped of the protective class rights Article 5(4)(b) gives it. What stands is a design rule binding this instrument and anyone amending it: economic participation is the maximum, and no control right, veto or governance privilege attaches to the Reserve's holdings (DC-32). It is a commitment the text keeps, not a lock the text can impose.

17. You are seizing equity in companies Europe does not govern

The objection, at full strength. The warrant obligation reaches undertakings incorporated in Delaware or Singapore, whose shares sit offshore and whose systems are built offshore, because their services are used in the Union. Public international law lets the Union regulate foreign conduct only where it has immediate, substantial and foreseeable effects in the internal market (Gencor T-102/96; Intel C-413/14 P), and even then it regulates conduct, not ownership: no effects-doctrine case has ever required a foreign parent to dilute its own capital. Third-country governments will treat a compulsory 3 % subscription in their champions as expropriation by regulation, answerable under investment treaties and trade commitments, and they will retaliate. The Union would be claiming a power it would never concede to others: a foreign statute demanding 3 % of a European champion's equity as the price of serving that market.

What it gets right. A warrant on a foreign parent whose only Union link is that its website resolves would overreach and would deserve to lose. The nexus must be economic substance in the Union, not accessibility. Retaliation is a real cost and reciprocity a real argument.

The answer the instrument must give. Four structural choices, all already in the articles. First, the trigger is Union commerce, not Union accessibility: designation requires provision in the internal market with EUR 7,5 billion of annual Union turnover in at least three Member States (Article 3(2)(a)), an economic-presence test far above any effects-doctrine threshold, and the rents being shared are by construction rents drawn from Union users. Second, the undertaking is the group: 'undertaking' consolidates linked enterprises (Article 2(1)), the single-economic-unit principle of Union competition law (Akzo Nobel C-97/08 P), so no thin Union subsidiary can shield the parent to which the automated services' value actually accrues. Third, the mechanism respects foreign company law rather than purporting to override it: for undertakings governed by third-country law the subscription is an obligation of result (Article 5(5)), enforced through Article 13 as a condition of continuing access to the internal market, the architecture of every market-access condition the Union already imposes, and the company-law derogations of Article 5(7) reach Member State law only. Fourth, the condition is universal: Union undertakings bear it identically, so a treaty claim or trade panel has no discrimination to hold on to, and reciprocity runs in the instrument's favour, because a Union that asserts the principle accepts it from others.

On the practical incidence the objection now has a concrete answer rather than an abstract one. Under the designation test of Article 3(2)(b) as amended, the undertakings presumptively designated on August 2026 figures span the United States, Taiwan and, near the threshold, the Union itself (evidence/designation-count.md), so the measure's first-years incidence is no longer a set consisting exclusively of undertakings of one third country, and the universality of the condition is visible in the designated set itself rather than asserted about a hypothetical one.

Design consequence. DC-10's market-access nexus and DC-2's group consolidation are this objection's answers in law. It adds one rule of its own: for undertakings governed by third-country law, the warrant is an obligation of result enforced as a market-access condition, never a purported override of foreign company law (DC-33).

18. Corporate structuring will simply route around all of this

The objection, at full strength. Every mechanism in this Regulation assumes a moment at which value becomes visible, and corporate law exists to move that moment. A hyper-automated undertaking is the ideal escape artist: it needs little capital, so it need never list; it generates cash, so it can pay its founders through decades of dividends, leveraged recapitalisations and selective buybacks while the warrant sits dormant for ever. If it must eventually deal, it deals in a capital stack the Reserve does not sit in: seed through Series G preferred with two- and three-times participating liquidation preferences, so that a sale distributes everything to preferred holders and the Reserve's three per cent of common is worth precisely nothing. If the trigger still nears, the group demerges: model weights, training infrastructure and user base migrate to an unlisted sister, and what floats is a low-margin European storefront. If all else fails, a friendly private credit fund converts debt to equity in a preventive restructuring, old equity is extinguished by court order, and the same people own the same models through a new company the following morning. Meanwhile the shares the Reserve does hold in third-country parents are taxed at source at rates it cannot reclaim, because a supranational body resident nowhere has no treaty to invoke. Each of these is ordinary practice, not abuse; the instrument is a crystallisation event in a world that has spent forty years learning to avoid events.

What it gets right. All of it. An instrument that hangs on a single trigger will be routed around by people who structure for a living, and a regulation that says so only in its recitals has conceded the point. The answers below are additions the file did not have; they came from outside review and the memorandum says so.

The answer the instrument must give. Close the timing, the ranking, the perimeter and the exit, and be honest where the closure is partial.

Timing first: the warrant no longer waits on a sale. Article 5(3) makes it crystallise where shareholder extraction over three consecutive years exceeds 25 % of the undertaking's turnover from the covered activity, which is the point at which staying private has become a way of paying oneself rather than a way of building; and in any event on the seventh anniversary of issuance, whatever the undertaking has or has not done. An undertaking may still stay private for ever. It may no longer stay private and be paid.

Ranking second: Article 5(4)(b) entitles the Reserve to shares ranking, for dividends and for the proceeds of any sale or liquidation, equally with the most favourably ranking class created after designation, and with ordinary shares otherwise. Three per cent of a common tranche standing behind a three-times participating preference is not three per cent of anything, and the instrument now says which three per cent it means, without taking from investors the preference they actually paid for before designation. Article 5(10) makes any reorganisation whose main purpose or effect is to put the Reserve below where that paragraph places it ineffective as against the Reserve, while leaving it good between the parties to it.

Perimeter third: Article 5(11) requires the transferee to issue a warrant of its own where automated assets go to an affiliate or a commonly controlled entity or leave at less than arm's length, and leaves the transferor bound in respect of what it retains. The aggregate across the covered undertaking and every transferee is capped at the stated percentage of their combined capital, so that closing this route cannot turn into a multiplier on a group that divides itself. Article 2(14) names what may not be walked out of the door: model parameters, training and inference infrastructure, data sets, the intellectual property the service depends on. This is the succession logic of merger control rather than an invention.

Exit fourth: Article 5(12) reattaches the obligation where the automated assets emerge from insolvency or a restructuring under Directive (EU) 2019/1023 into an entity controlled by the same people, or in which those people hold the majority of the economic rights, and requires a warrant to be issued afresh within three months. A genuine failure still extinguishes the Reserve's holding, as it extinguishes every shareholder's, because the Reserve is an owner and owners bear that. What it does not do is bless the version where the owners survive and only the obligation dies.

And the honest partial: source taxation. A Union regulation cannot confer on the Reserve a treaty benefit that a third country has not granted, and a body resident nowhere may be withheld against at the full statutory rate. Two answers were drafted and struck. Instructing the Commission to negotiate with third countries would have commanded a prerogative that Article 218 TFEU places elsewhere. Permitting the Reserve to route its holdings so as to reduce the rate would have written treaty shopping into an instrument whose entire case is that capital should be broadly owned and should pay what it owes; hostile counsel had the headline ready, and was right to have it. What survives is Article 8(5): the Reserve publishes each year the tax it could not reclaim, and the Commission answers for it in the Article 14 report. That is a smaller answer than the problem, and it is the honest one. A leak reported annually is a leak the campaign can be judged on; a leak nobody measures only grows.

One route counsel kept and the instrument deliberately leaves open. An undertaking may borrow heavily from lenders who are nobody's relation, pay them market interest, and arrive at the long-stop worth less than it would have been unlevered. That is not extraction and the definition should not pretend otherwise: the money goes to strangers, not to insiders, and the founders are poorer by exactly the same proportion as the Reserve. Every shareholder in a leveraged company owns a smaller claim on a larger balance sheet, and an instrument that took 3 % of the equity cannot also insist the equity be unencumbered. What the file must not do is confuse that with the routes above, where value leaves for pockets that are the same pockets. Leverage is a risk the Reserve takes as an owner; extraction is a transfer the Regulation stops.

Design consequence. DC-35: crystallisation triggers on extraction and on time, not only on a sale. DC-36: subject to the rescue carve-out, the Reserve's shares rank with the most favoured class created after the designation, leaving earlier preferences untouched, and subordinating arrangements are ineffective as against the Reserve. DC-37: the transferee issues its own warrant over its own capital; the transferor stays bound for the automated assets it retains, and the aggregate across them never exceeds the stated percentage of their combined capital. DC-38: unrecoverable source taxation is published annually; the instrument does not direct the Reserve to arrange its holdings to reduce it. DC-39: the obligation reattaches where the same owners reacquire the assets out of a restructuring.

19. Sweden tried this and could not even legislate it

The objection, at full strength. This mechanism is not new and it has already failed, in the one country most likely to have carried it. The Meidner plan, endorsed by the Swedish trade union confederation's congress in 1976, required profitable undertakings to issue new shares worth 20 % of their annual profits to collectively held funds, year after year, until those funds held a majority of the shares. That is this instrument's mechanism: a compulsory issuance of new equity to a collective holder, settled in shares rather than cash, calibrated to the undertaking's own prosperity.

It did not survive its own drafting. By the time the Riksdag legislated in 1983, the compulsory share issuance had been abandoned and replaced with a levy on profits and on the wage sum, capped in aggregate and limited to holdings of no more than 8 % of any one undertaking. The design converted, under political pressure and before enactment, into precisely the thing this memorandum spends Article 8 and objection 2 insisting this instrument is not: a tax. Then Swedish business mobilised against even the diluted version on a scale without modern parallel, with a demonstration on 4 October 1983 that drew between 75 000 and 100 000 people where the organisers had expected 5 000, and repeated it annually until a change of government abolished the funds in 1991, against minimal resistance from the movement that had proposed them.

The reading is unkind and it is available to anyone who knows the file. In the most union-friendly advanced economy in the world, at the peak of organised labour's power, with a sympathetic government, the compulsory issuance of equity to a collective fund proved unlegislatable, converted into a fiscal measure at the drafting stage, and was repealed within eight years by the first government that wanted to. A drafter who does not know this has not done the reading. A drafter who knows it and omits it is hiding it.

What it gets right. Almost all of it, as history. The mechanism is the same family. The conversion to a levy is real and is the sharpest available evidence for objection 2's characterisation attack, because it shows experienced legislators reaching for the fiscal instrument when the equity instrument became politically impossible. The 1991 abolition is a genuine instance of the raid this instrument drafts Chapter VI against, and it belongs beside Poland, Spain and Ireland in the evidence base rather than being left out because it is inconvenient. The mobilisation is a real prediction about what happens when capital reads a proposal of this kind as a transfer of control.

The answer the instrument must give. Three differences, each of which maps onto a specific reason the Swedish design failed, and none of which is cosmetic.

First, terminus. Meidner's funds were designed to accumulate without limit until ownership changed hands, and its author said so; that was the point, not a side effect. This instrument takes 3 % of fully diluted capital once, per designation, with the aggregate across an undertaking and every transferee capped at that same percentage under Article 5(11) and Article 7(2). There is no path from this instrument to control, because the arithmetic does not permit one. An opponent who says otherwise is arguing against a different proposal.

Second, control. The Swedish funds voted, and they were run by the unions. The scholarship on the defeat is consistent that the mobilisation was about who would govern the companies rather than about the money, and the 8 % ceiling was itself an attempt to answer that fear. The Reserve holds non-voting shares under Article 5(4)(a), is forbidden by Article 9 from seeking or exercising influence, and DC-32 makes economic participation the maximum. The constituency that made 1983 possible was employers who believed they were being nationalised in instalments. That belief cannot be formed about an instrument that cannot vote.

Third, beneficiary. The Swedish funds accrued to a labour movement, which made them a party question in a two-party fight. This instrument's distributions accrue equally to every citizen of the Union under Article 10, including the shareholders and the employees of the designated undertakings themselves. That does not make opposition impossible. It makes "them against us" harder to draw.

What this file will not claim is that the differences make the political economy safe. They do not. The honest position is that Sweden shows the mechanism can be legislated only if it is visibly bounded, visibly powerless as to control, and visibly universal in who it pays, and that a proposal failing any of those three has a documented way of dying. This instrument is drafted to satisfy all three, and objection 4 already prices the residual political risk rather than denying it.

Design consequence. DC-41: the instrument must be bounded, without votes or control rights, and universal in who it pays, all three at once, because the Swedish precedent shows that failing any one of them is sufficient to lose. The word to avoid is passive. The Reserve does assert claims: Article 5(4)(b) fixes its rank and Article 5(10) makes subordinating arrangements ineffective against it, and an opponent will call that anything but passive. The defensible claim is narrower: the Reserve holds no votes, appoints no one, and has no say in the management of the undertaking. Even that is not the end of it. Hostile counsel makes the further point that a rank the undertaking cannot subordinate constrains how it raises distressed or preference capital, since new money normally demands seniority, and calls that a veto over capital structure in substance. It is not a governance right and Article 5(10) leaves the arrangement effective between the parties to it. It was, all the same, a constraint on financing that the word passive would conceal, and it is now answered in the text rather than carried: the second subparagraph of Article 5(10) lets genuinely new money rank ahead of the Reserve, at arm's length, from persons unconnected with the undertaking and its controllers, while the undertaking is in a likelihood of insolvency or must meet a prudential requirement, and only to the extent of the new consideration provided. Rescuers are paid before the Reserve; engineered seniority beyond the new money is still caught, because that is where the abuse lives, and the undertaking bears the burden of proof. DC-42: the wage- earner funds belong in the evidence base as a raid precedent and in this file as an objection, stated before an opponent states it.

20. Taking the voteless shares hands control to the people you named as the problem

The objection, at full strength. The instrument takes three per cent of the economic value and none of the votes. The arithmetic of that is not neutral. Before the warrant, the existing holders own all of the capital and cast all of the votes; after it they own ninety-seven per cent of the capital and still cast all of the votes. Their control per euro of their own money has gone up, by about three per cent, and the Union has handed it to them. A file whose diagnosis is that the ownership of productive capital is too narrow responds by creating the largest permanently voteless shareholder in Europe and widening, at every covered undertaking, precisely the wedge between cash-flow rights and control rights that corporate-governance scholarship identifies as the most reliable predictor of minority expropriation and unaccountable management. The founders of the frontier firms already hold control through dual-class structures on minority economic stakes. This gives them more of it, for free, by statute.

The legitimation point is worse than the arithmetic. Once the Union itself accepts voteless equity as the form its own citizens' participation takes, no Union institution can coherently object to dual-class listings, loyalty shares or any other separation of ownership from control. The instrument does not merely tolerate the wedge; it endorses it, in a Regulation, on behalf of 450 million people.

What it gets right. The mechanical claim is exactly true and this file should not pretend otherwise: non-voting shares raise the relative voting weight of every voting share, and the raise here is real. If the diagnosis were about who controls firms rather than about who receives the returns to capital, the instrument would be answering a question it did not ask, and answering it in the wrong direction.

The answer the instrument must give. Not that this is the wrong lens. An earlier draft of this passage called the governance reading a category error, on the ground that the instrument's claim is distributional. Four review gates rejected that in the same terms, and they were right: an instrument that alters the ratio of cash-flow rights to control rights at every undertaking it touches is a governance instrument in effect, whatever it is in intention, and it is properly judged on its effects. The answer below concedes the effect and defends the choice.

The choice was between three positions and the file has taken the least bad one. A Reserve with votes proportionate to its stake is a politically directed shareholding in every frontier firm in the internal market: it is objection 9's honeypot with a lever attached, and objection 16's golden share, which the Court has struck down in every form it has taken since Commission v Germany (C-112/05). A Reserve with no stake at all abandons the distributional claim entirely. A Reserve with economic participation and no control accepts a real governance cost in exchange for the only position that survives Article 63 and the golden-share line. Article 5(4)(a) and Article 9(1)(a) to (c) put that beyond the drafter's later convenience, and DC-43 keeps any successor honest about it.

On magnitude, one comparison and its limits. Three per cent of enlarged capital raises the remaining holders' voting weight by roughly one thirty-second of itself, and buy-back programmes of that order run routinely at the undertakings this instrument covers without anyone calling them a transfer of control. That comparison is offered for size and for nothing else, and it does not survive being pushed further: a buy-back is elective, is executed for consideration and returns cash to the holders whose weight rises, whereas this is compulsory, is subscribed at nominal value and returns nothing to them. Reviewers were right to say the two are not alike in kind. They are alike in the one respect the paragraph uses them for, which is how much the voting weight of a remaining share actually moves.

On legitimation, the objection lands and is only partly answered. Voteless shares are not this instrument's invention: Union company law has permitted them for decades and every Member State provides for them, so the Reserve takes a class the market already trades. But the Union has lately been legislating in the other direction, adding safeguards around multiple-vote structures precisely because the separation of ownership from control is understood to be a hazard, and a Regulation that plants a permanent voteless block in every covered undertaking sits uneasily beside that. The honest statement is that the instrument spends some of the Union's authority on the proposition that economic participation without control is a legitimate form of ownership, and that the price is real. It is paid deliberately, because the alternative is a Union holding votes in the undertakings it regulates, and this file would rather defend a wedge than defend that.

The precise interaction with the Union's multiple-vote share legislation is recorded as an open acquis interface point rather than argued here from a citation this file has not yet verified, and it joins the acquis interface points already recorded on the list before filing.

Design consequence. The instrument's claim is distributional and no amendment gives the Reserve votes without arguing itself as a different instrument, but the governance effect is conceded and defended rather than disclaimed (DC-43). The relative voting effect on existing holders is stated with its size, and the buy-back comparison is confined to magnitude with its disanalogy stated in the same breath (DC-44).

21. Why three per cent, and not one, or ten

The objection, at full strength. The number is asserted. Nowhere in this file is it derived. Article 52(1) of the Charter permits a limitation on the right to property only where it is necessary and genuinely meets an objective of general interest, and necessity is the question a bare figure cannot answer: if three per cent achieves the objective, one per cent is the less intrusive means and the measure fails; if one per cent does not achieve it, the file has not shown why three does. The reviewing court will ask the legislature to show its work, as it asked in every proportionality case worth citing, and this file will hand it a round number that sits comfortably below the level at which the objection "you are nationalising them" becomes easy to make. That is a political calibration wearing the clothes of a legal one, and the Commission's Legal Service will see it on the first reading. Every other number in the instrument is tied to something: the thresholds in Article 3 to measurable quantities, the seven-year long stop to the observed interval between designation and realisation. The one number that determines how much is taken is tied to nothing.

What it gets right. All of it. This is the weakest load-bearing point in the file, and no quality of drafting anywhere else repairs it. It would be worse than useless to answer it here with a derivation invented to fit a figure already chosen: that is precisely the vice the objection names, and a reviewer who caught it would be entitled to discount everything else in this memorandum.

The answer the instrument must give. The objection is correct, and the attempt to answer it produced something worse than a missing derivation: a demonstration that no derivation of the kind a court would want can be produced at all while Article 1 stands as drafted.

The published model is linear in the percentage. Annex II's arithmetic, as implemented in the simulator on the campaign site, accumulates the Reserve's capital by adding the warrant percentage of each crystallising flow, and every distribution downstream is a fraction of that capital. Doubling the percentage doubles the dividend at every horizon and halving it halves it, exactly, in every scenario the simulator offers. The computation is set out in evidence/warrant-percentage.md and anyone can rerun it. There is no threshold, no kink and no discontinuity anywhere on the curve, which means there is no percentage at which the instrument starts working and below which it does not.

That is fatal to the obvious defence. Necessity under Article 52(1) asks whether a less intrusive measure would achieve the objective. If the objective is Article 1's, the participation of citizens in the capital value created by hyper-automated undertakings, then one per cent achieves it, and so does one tenth of one per cent, because each produces participation and the Article states no quantity that participation must reach. An opponent does not even need to argue that three per cent is too much. They need only observe that a smaller figure achieves the stated objective, and the necessity limb fails on the instrument's own words.

So the defect is not in this memorandum, and it cannot be repaired here. It is in Article 1. An instrument that interferes with Article 17 of the Charter and states its objective qualitatively has left the necessity limb with nothing to be tested against, and no amount of argument about three per cent substitutes for that. Either the objective acquires a quantity against which a percentage can be measured, or the Article 52(1) defence rests on the proportionality limb alone, which asks only whether the burden is excessive and to which the answer is the ceiling rather than the floor.

The ceiling is where the file's defence honestly sits today. A percentage large enough to strip existing holders of the substance of their property crosses from a regulation of use into a deprivation. Hauer (44/79) is the authority for the distinction itself and for the proposition that the right to property may be regulated in the general interest; it concerned a temporary restriction on planting vines and is not authority for a compulsory transfer of equity, and this file does not offer it as one. James and Others v United Kingdom (1986) is cited for the same distinction and for nothing further: the taking there was accompanied by compensation, so it is authority about where the line runs and not about what may be taken without paying for it. Objection 1 carries that argument and its honest residue, which is that no decided case places a permanent, non-crisis, uncompensated dilution of a healthy undertaking on the right side of the line. Three per cent is comfortably below any plausible deprivation ceiling. Being below a ceiling is a proportionality argument, not a necessity argument, and this file will not dress one as the other.

That repair has now been made, and this passage records both the repair and its honest limit. Article 1(2) states the objective as an ownership position: holdings per citizen of the order of six months of the median equivalised disposable income in the Union, in constant prices, within a generation of the first designations. On the published model and the amended designation test of Article 3(2)(b), the aggregate designated value is of the order of EUR 20 trillion on August 2026 figures (evidence/designation-count.md), and the stake per adult within a generation is about EUR 8 400 at three per cent, EUR 5 600 at two and EUR 2 800 at one (evidence/sizing-the-ask.md). Six months of median equivalised disposable income is of the order of EUR 9 000. Three per cent is therefore the smallest integer percentage capable of approaching the stated objective, one per cent manifestly cannot, and the necessity limb of Article 52(1) has, for the first time, a quantity to test against, published, reproducible, and revisited at every Article 14 report, where the same clause that could raise the percentage on the evidence is the clause that lowers it.

The honest limit is the circularity a reviewer will point out and this file states first: with a linear model, any target-percentage pair is arguable, so the target itself is not derived, it is chosen, openly, in the enacting terms. That is not a defect the drafting can remove; it is what a legislative objective is. The case law requires a stated objective and a reasoned relationship between the measure and it, not an objective that derives itself, and no instrument's does. What this file no longer does is assert a percentage against no objective at all.

Design consequence. The percentage is derived as the minimum integer consistent with the objective stated in Article 1(2), on a published and reproducible model, and the circularity inherent in choosing the target is stated rather than concealed (DC-45). Article 1(2) carries the quantified objective, assessed in each Article 14 report in both directions, and the derivation is revisited whenever the designated set or the model moves (DC-46).

The constraints table

The articles are drafted against this table. A draft that violates a DC fails review regardless of its prose.

DC	Constraint	Source objection
DC-1	Prospective warrants on future value; never retroactive transfer	1
DC-2	High, objective, group-consolidated thresholds	1, 5, 17
DC-3	Statutory passivity; crystallisation only at defined statutory events, never at a discretionary or political one	1, 6
DC-4	No monetary flow from undertakings; instruments only	2, 4
DC-5	Distributions are property income of the reserve	2
DC-6	Severable layering for partial ECI registration	2
DC-7	Union warrant standards; Member State custody via pension rails	3, 14
DC-8	Minimum standards, not uniform machinery	3
DC-9	Obligation never convertible into a flow charge	4
DC-10	Market-access nexus, not establishment nexus	5, 17
DC-11	Substance-over-form headcount consolidation	5
DC-12	Event-triggered crystallisation; no ongoing valuation	6
DC-13	Permanently non-voting economic interests	6, 16
DC-14	Dividend communicated as compounding from small	6
DC-15	Recitals argue ownership and mechanism, never wage-share decline	7
DC-16	Value stated under both futures	7
DC-17	Strongest nulls cited in the memorandum itself	8
DC-18	Stated falsification condition with a date	8
DC-19	No sovereign debt holdings	9
DC-20	Entrenchment from day one: express amendment only, published independent assessment, honesty about Treaty limits	9
DC-21	No emergency clause	9
DC-22	Individual claims incapable of surrender or seizure; distributed amounts owned and inheritable	9, 13
DC-23	Raid resistance claimed as friction, never impossibility	9
DC-24	Self-executing by operation of law; no programme machinery	10
DC-25	Consumer surplus affirmed; instrument additive to it	11
DC-26	Explicit scope-and-limits recital	12
DC-27	UBI distinction carried by legal form	13
DC-28	Funding-source novelty only; existing rails for delivery	14
DC-29	Interference capped at the stated percentage in execution; independent, separately challengeable valuation; judicial review	1
DC-30	Essential elements (trigger, reserve ownership, entitlement, interference) in the articles, never delegated	1
DC-31	Wherever a per-citizen payout figure is shown, the per-citizen stake in the Reserve is shown beside it	15
DC-32	Economic participation is the maximum: no control right, veto or governance privilege attaches to the Reserve's holdings, binding this instrument and anyone amending it, without pretending to bind a future legislature	16
DC-33	For third-country-law undertakings the warrant is an obligation of result as a market-access condition, never an override of foreign company law	17
DC-34	No tier between valuation and the courts; correction is ex post and the transaction never waits	6
DC-35	Crystallisation triggers on shareholder extraction and on time, not only on a sale	18
DC-36	The Reserve's shares rank with the most favoured class created after the designation, leaving earlier preferences untouched; subordinating arrangements are ineffective against the Reserve, save for new money at arm's length from unconnected persons in rescue, to the extent of the new consideration	18
DC-37	The transferee issues its own warrant over its own capital; the transferor stays bound for what it retains, and the aggregate across them never exceeds the stated percentage of their combined capital	18
DC-38	Unrecoverable source taxation published annually; the Reserve is not directed to arrange its holdings to reduce it	18
DC-39	The obligation reattaches where the same owners reacquire the assets out of a restructuring	18
DC-40	Below a de minimis the distribution interval lengthens; it never becomes a reason not to pay	15
DC-41	Bounded, without votes or control rights, and universal in who it pays: the Swedish wage-earner funds show that failing any one of the three is enough to lose. The Reserve does assert claims as to rank and against subordination; what it never acquires is votes, board seats or any say in management	19
DC-42	The wage-earner funds are stated as an objection here and as a raid precedent in the evidence base, before an opponent states them	19
DC-43	The claim is distributional and no amendment gives the Reserve votes without arguing itself as a different instrument; the governance effect on remaining holders is conceded and defended, never disclaimed as the wrong lens	20
DC-44	The relative voting effect on existing holders is stated with its size, not left for an opponent to compute, and any comparison drawn to buy-backs is confined to magnitude with its disanalogy stated alongside	20
DC-45	The percentage is the minimum integer consistent with the objective stated in Article 1(2), derived on a published, reproducible model; the circularity of choosing the target is stated, never concealed	21
DC-46	Article 1(2) carries the quantified objective, assessed in each Article 14 report in both directions; the derivation is revisited whenever the designated set or the model moves	21

Status

Objections are OPEN until the articles answer them, save where noted below; 12 is CONCEDED by scope. Objection 18 was added on 21 August 2026 from an external review of corporate-structuring routes around the instrument. Unlike objections 16 and 17, it did not find the answers already in the articles: it describes five routes. Four of them are now closed by amendments to Articles 2 and 5. The fifth, source taxation, prompted an amendment to Article 8 but is not closed, and the answer to that objection says so: publication is a smaller answer than the problem. A sixth item, borrowing at arm's length, was considered and deliberately left open, because it is not the extraction the objection is about. It is the first objection in this file whose answer had to be written rather than cited.

Objections 16 and 17 were added on 20 August 2026 from an external challenge (the Article 63 golden-share line; qualified-effects overreach); their answers were already in Articles 3, 5(5) and 9, which is what drafting against the table is for, and the residue they add is DC-32 and DC-33. Objection 1 carries the largest legal risk and objection 6 the largest design risk. Gate 1's admissibility letter still leads with objections 1 and 2, but recalibrated by the drafting research: registration is the lower hurdle (manifestly-outside test, partial registration, the registered wealth-tax ECI), so the letter tests the characterisation for the Council stage, not for the register.

Objection 2's fiscal reading has now been reached independently by three readers: hostile counsel on 27 August 2026, the ECI Forum's advisers on 27 August 2026 and an ECI veteran on 5 September 2026 (pipeline/EXTERNAL-REVIEWS.md, review 5). The objection is unchanged; what changes is its status, from an adversary's argument to the ordinary reader's first impression, which the presentation, not the drafting, must now answer.